Pre-mix cake recipes have been available for a while at grocery stores. But few know that when such product came out, and however good cakes could be, it WASN’T a commercial success. It only became one the day users also needed to crack eggs to make a cake, not only add water or milk. That’s because suddenly, it turned the result into a home-made cake, versus a pre-made cake.
In my previous TIC blog, I talked about how critical it is to consider AI skills and human skills applied in concert in today's rapidly evolving technological landscape. The above analogy teaches us one essential point about it, versus thinking only about AI through taking over entire tasks or roles. As AI continues to advance, it is crucial for people managers, executive decision-makers, and frankly whole organizations to develop both technical and human skills to effectively collaborate with, and fully leverage AI skills.
As explained in that previous blog, TI leaders have a strategic responsibility in this endeavor. They must be pivotal in helping their organization practically make it a reality. Specifically, it starts with rethinking the aging skill-mapping exercise, and turn it into one that integrates both AI and human skills.
That’s for the theory. How could ambitious TI leaders practically act upon that idea? Here is an attempt to grasp the elements for such a framework.
Human Skills, AI Skills, and “Shared Competencies”
For human skills, it is possible to think of skills in terms of a two-dimensional matrix. The first dimension is the type of skill, such as analytical skills, problem-solving skills, communication skills, and so on. The second dimension is the level of skill, which can be measured by proficiency, experience, or education. This is historically how human skills have been categorized. On the other hand, AI skills require a different approach. AI skills can be divided into different categories such as machine learning, natural language processing, computer vision, robotics, and so on. The dimensions for these skills may vary depending on the specific AI application or domain.
Yet there are also what could be called “Shared Competencies” — For example, critical thinking, AI-assisted creativity, and human-machine collaboration are important skills for both AI and human intelligence. These competencies may be assessed using different parameters, but they can be used to measure the effectiveness of both human and AI systems. That said, it remains challenging to create a form of “unit of skill” that would be common to AI skills and human skills, though. This is because AI skills and human skills operate differently and require different types of assessment and training.
Core Attributes Leaning on the Human Side of Skills
A mapping and categorization for the above 3 skill categories must be understood as a multi-dimensional matrix. Dimensions, categories and attributes of that matrix are naturally dynamic. An initial broad definition and structuring of core attributes is therefore recommended. For a base, I suggest considering:
- Technical AI Skills: These are the foundational skills that are required for AI development and implementation, such as programming languages, algorithms, data structures and literacy, machine learning frameworks, statistical models, and so on. This dimension would also include skills related to machine learning, natural language processing, computer vision, data analysis, and other AI-specific technologies.
- Human-Centered AI Skills: This dimension would include skills related to human-AI interaction, such as user experience design, ethical considerations, empathy, and cognitive psychology.
- Industry-Specific Skills: This dimension would include skills specific to a particular industry, such as healthcare, finance, education, or manufacturing, that could be enhanced by AI.
- Core Human Skills: This dimension would include skills that are foundational for any profession and essential for effective collaboration with AI, such as communication, critical thinking, creativity, problem-solving, and adaptability.
Primary Dimensions for AI-Skills and “Shared Competencies”
AI skills don’t necessarily have to be categorized upon purely technical dimensions. For example, considering AI skills related to generative AI as a subset of Machine Learning. In fact, for simplification purposes, I recommend looking at the dimensions through easily relatable human skills. Here is an attempt for an early list of key dimensions of AI skills with likely crossover with Human competencies:
- Problem-solving: AI is used to solve complex problems, and AI professionals need to be skilled in critical thinking, problem identification, and problem-solving.
- Perception: this dimension involves the ability of AI systems to perceive and interpret the world through different types of sensors and data sources. Experimented in a business context: Retail stores using computer vision to track customer behavior and analyze traffic patterns.
- Reasoning: this dimension involves the ability of AI systems to use logic and reasoning to solve problems and make decisions
- Knowledge representation: this dimension involves the ability of AI systems to represent and store knowledge in a structured way.
- Learning: this dimension involves for example the ability of AI systems to learn from data and improve over time.
- Ethics: this dimension relates to the ability to make ethical and responsible decisions, taking into account social, cultural, and environmental factors. While AI can make objective and data-driven decisions, it lacks the ability to account for ethical considerations that require human judgment and empathy. Human ethical reasoning and values can be a valuable multiplier to AI skills in areas such as healthcare, law, and finance.
- Adaptability: this dimension involves the ability to learn and adjust to changing circumstances and contexts. While AI can be trained to perform specific tasks, it lacks the flexibility and agility of human beings to learn and adapt to new situations. Human adaptability can be a valuable multiplier to AI skills in areas such as customer service, logistics, and project management.
- Intuition: this dimension involves the ability to understand and act on unarticulated or implicit knowledge, such as emotions, motivations, and social dynamics. While AI can analyze vast amounts of data and patterns, it lacks the ability to intuitively understand human behaviors and contexts. Human intuition can be a valuable multiplier to AI skills in areas such as marketing, psychology, and education.
- Trustworthiness: this dimension relates to the ability to build and maintain trust with customers, partners, and stakeholders. While AI can provide objective and accurate information, it lacks the ability to establish and maintain human relationships and credibility. Human trustworthiness can be a valuable multiplier to AI skills in areas such as sales, leadership, and public relations.
This list is not exhaustive and can vary based on the specific role, industry, or organization involved. Therese dimensions are also not mutually exclusive, and there may be overlap or interaction between them. Additionally, the relative importance and value of each dimension may vary depending on the application and context of AI skills.
Making the Case for the New Extended Mapping
There are many job families and industries where the combination of AI and human skills can generate significantly better outcomes than using either one alone. But we could equally think of mediocre outcomes implied by the poor application of AI and human skill working in concert.
Picture a production operation (in manufacturing, content creation, etc.), and the AI system is programmed to optimize production speed and reduce costs, but lacks the ability to identify and address quality issues. The human workers lack technical training and struggle to operate the system, leading to frequent errors and production defects.
In a well-skilled situation, the AI system is trained in both speed and quality optimization, and can detect and address quality issues in real-time. The human workers are trained in technical skills and can effectively operate the system, minimizing errors and ensuring high-quality products.
As AI becomes better at performing tasks that were previously the sole domain of human workers, there may be a need to retrain workers or shift their roles to take advantage of the new opportunities created by AI. Similarly, as AI becomes more integrated into decision-making processes, it will be important to ensure that the decisions being made are ethical and align with the values of society.
Ideally, to avoid wrong outcomes and mitigate associated risks, I encourage thinking through use-cases where human skills, if combined with the AI skill, will act as value multiplier. In particular, for situations in which the balance between humans and AI skills will evolve and require monitoring. Amongst other situations:
- Decision Making and Planning — Think about predictive analytics and machine learning algorithms that can identify patterns and make recommendations based on data. Expertise in the domain or industry being analyzed, and the ability to make strategic decisions based on a combination of data and intuition is definitely where human skills can be used as value multiplier. But as decision making and planning tools become more sophisticated, there may be a shift towards greater reliance on automated recommendations, potentially leading to a loss of human expertise in decision making. It will be critical to monitor the AI/Human skill balance and ensure that human skills in strategic thinking and decision making are still valued and developed.
- Creativity and Imagination — Generative AI models that can create original artwork, music, or other media. Did someone said ChatGPT? Human skills acting as value multiplier lie in the ability to conceptualize and develop original ideas, and to apply creative thinking to problem-solving and innovation. Of course, as generative AI models become more advanced, there may be a risk of undervaluing the importance of human creativity and imagination. Natural language processing technologies become more advanced, but there may be a risk of over-reliance on the technology. This is why it will be massively important to monitor the AI/Human skill balance and ensure that human skills in creativity and innovation are not dismissed upon misunderstanding the true capabilities of AI systems.
- Social and Emotional Intelligence — AI-powered virtual assistants or chatbots that can detect and respond to emotions in human communication. Human Skill Value Multiplier: The ability to build and maintain relationships with customers and clients, and to provide emotional support and empathy when needed. Evolution and Monitoring: As AI-powered virtual assistants become more common in customer service roles, there may be a risk of undervaluing the importance of human social and emotional intelligence. It will be important to monitor this balance and ensure that human skills in relationship-building and emotional intelligence are still valued and developed.
Where to start?
A framework for monitoring the balance between AI and human skills in a workplace could include the following components:
- Identify key job roles and tasks: Start by identifying the job roles and tasks that are currently being performed by humans and those that are being automated with AI.
- Define the level of AI integration: Determine the level of AI integration required for each job role and task. This could include full automation, partial automation, or augmentation of human skills.
- Assess current skill sets: Assess the current skill sets of employees in each job role and task to determine their proficiency in both AI and human skills.
- Identify skill gaps: Identify any gaps between the current skill sets and the skills required for the desired level of AI integration. This could include identifying areas where employees may need training or upskilling to perform their jobs effectively in a changing work environment.
- Develop training and upskilling programs that are tailored to the specific needs of each job role and task. This could include training programs for both AI and human skills, as well as programs that focus on developing hybrid skills that combine AI and human capabilities.
- Measure and track progress: Establish metrics to measure and track progress in achieving the desired balance between AI and human skills. This could include measuring employee proficiency in AI and human skills, as well as monitoring the effectiveness of training and upskilling programs.
- Continuously assess and adjust: Continuously assess and adjust the framework based on the changing needs of the organization and the evolving capabilities of AI technology. This could include regularly reviewing job roles and tasks to identify new opportunities for AI integration, as well as adapting training and upskilling programs to keep pace with technological advancements.
These points will resonate with TI leaders. It is not surprising as similar models are commonly applied with human skills. Today’s challenge is to realistically integrate these action points with AI skills. Overall, any framework integrating AI skills in the new skill mapping exercise should focus on creating a culture of continuous learning and adaptation to ensure that the organization is well-positioned to take advantage of the benefits of both AI and human skills.
Of course the training programs for AI and human skills would differ as they require different types of expertise and knowledge. However, there could be some overlap or complementary aspects in the training programs that could enhance the combination of AI and human skills.